Introduction

The kelp Saccharina latissima (Linneaus) (Laminariales, Phaeophyceae) is widely distributed in temperate to polar areas of the northern hemisphere where it forms key habitats in the coastal zone (Kain 1979; Steneck et al. 2002; Bartsch et al. 2008). Kelp forests are among the most productive ecosystems on the globe, and the primary production enters the food web through grazing, exudation of dissolved organic carbon and as detritus (Mann 1973; Duggins et al. 1989). The large size of kelps and the extension of kelp habitats also give kelps an important structuring role as substrate for sessile organisms as well as shelter and nursery area, thereby stimulating a high biodiversity (Dayton 1985; Bruno and Bertness 2001; Lippert et al. 2001; Steneck et al. 2002). Kelps such as S. latissima exhibit a pronounced seasonal pattern of growth and decay, which influences the turnover of organic matter and the structure of the kelp habitat, thereby constituting the important role of kelps in coastal ecosystems (e.g., Krumhansl and Scheibling 2011).

New kelp tissue is initiated in the basal meristematic part of the blade just above the stipe, while old tissue is lost by erosion and decay from the distal part of the blade. In general, length growth of S. latissima is initiated in winter, peaks in spring and is reduced during summer and autumn (Kain 1979; Lüning 1979; Sjötun 1993). The seasonal pattern of blade erosion is roughly opposite to that of length growth as it peaks in July to December (Parke 1948). Seasonal fluctuations of light and ambient nutrient levels are believed to be major drivers of the growth patterns of Laminariales. At high nutrient levels and low light levels during late autumn and winter, Laminariales build up reserves of nitrogen (N) which are utilized to initiate growth in early spring when light levels increase (Chapman and Craige 1977). Rapid length growth can be sustained until late spring/early summer, after which it ceases. Instead of supporting further length growth during summer, photosynthesis products in this period are being stored for use in winter respiration and support of length growth during late winter/early spring (Black 1950; Bartsch et al. 2008). Hence, the tissue content of these carbon-based storage products peaks in late summer and declines to a minimum in late winter/early spring (Black 1950). Temperature, exerting key control on all metabolic processes, also markedly affects the production and respiration of S. latissima (Davison et al. 1991) and thus the seasonal growth cycle. Most studies have though focused on light and nutrient levels in explaining seasonal growth patterns and paid less attention to water temperature.

The importance of light and temperature for regulation of seasonal growth suggests that growth patterns may respond predictably to large-scale gradients in light and temperature across the species’ distribution range unless variation in local factors such as nutrient levels, salinity, exposure or water depth have overriding importance and disrupts such potential large-scale relationships. Year-round availability of N in the water has been shown to suppress the storage of carbon (C) and inorganic N in Laminaria longicruris and hence making length growth completely coupled to the seasonality of light (Gagne et al. 1982). Low salinity may suppress growth (Spurkland and Iken 2011), and differences in wave exposure and water depth may cause further variation in growth dynamics in kelps (Mann 1973; Gerard and Mann 1979; Vadas et al. 2004). In addition, the age structure of populations is likely to influence the dynamics (Parke 1948; Mann 1972; Lüning 1979; Druehl et al. 1987; Sjötun 1993).

Existing studies on growth dynamics in Laminariales often lack combined information on growth, loss, tissue composition, age structure and environmental conditions over entire annual cycles which limits the possibility of identifying how these aspects are related and whether they differ among sites. Furthermore, the present knowledge is still restricted to few places relative to the extensive distribution of S. latissima, and the vast majority of studies represent locations with high salinities which are optimal for growth of the species (27–33 psu, Gerard et al. 1987). Hence, detailed investigations describing and interpreting the ecophysiology of growth, seasonal growth and loss patterns in relation to thallus morphology, storage products, age structure, depth and physicochemical conditions of the location are needed. This paper presents a study of seasonal growth patterns in S. latissima, integrating these parameters. The aim was (1) to describe the growth and biomass dynamics of an established S. latissima population in a Danish embayment in terms of seasonal growth and loss, age structure, morphology, dry weight (DW) and C and N dynamics, (2) to explore how dynamics of growth, loss and storage products vary with water depth, (3) to discuss the findings in relation to the physicochemical conditions of the locality and (4) to relate the observed growth dynamics to previous studies from different latitudes with the perspective of identifying large-scale patterns in growth dynamics.

Materials and methods

Study site

The study was conducted in Aarhus Bay, Denmark (56°03.793N, 10°16.148E), in the period March 1999–March 2000. Aarhus Bay is an embayment situated at the Baltic Sea–Kattegat transition where low- and high-saline waters mixes. The bay constitutes an area of roughly 610 km2. The study area was restricted to an area of approximately 100 m in length where S. latissima was present and may not be representative for the entire bay. The seabed at the sampling site was composed of sand with scattered stones. During the period of the survey, six sampling and growth measuring events of S. latissima were performed at 7 m of depth. In July and September, additional measurements were conducted at 4 and 11 m of depth.

Physicochemical variables

Information on temperature, salinity, nutrient concentration and irradiance/light attenuation (Secchi depth) was obtained from the Danish National Aquatic Monitoring and Assessment Program (DNAMAP) that has monitored these variables 4–6 times per month according to common guidelines. Data were extracted from the national database on marine data (MADS) located at the Danish Centre for Environment and Energy (DCE) at Aarhus University. Values of temperature, salinity and nutrient concentration were based on measurements between 5 and 10 m depth. The light attenuation coefficient (K d) was calculated from measurements of the Secchi depth from the equation: K d = 2.3/Secchi depth. Irradiance levels at different depths were then calculated from the Lambert–Beer equation (e.g., Dennison et al. 1993): I z  = I 0 exp[−K d z], where I 0 is the surface irradiance, K d is the light attenuation coefficient, z is the depth, and I z is the light at a given depth.

Growth measurements

Divers collected 81 individuals of S. latissima from 7 m of depth on March 3, 1999. Blade length of the individuals ranged from 21 to 199 cm, with an average of 82 cm. All specimens were brought to the beach and individually tagged and measured in terms of length and width of the blade (width was measured at a distance of 10, 40, 70, 100, 130 and 160 cm from the transition of stipe to blade). The blades were kept humid by adding seawater during the measurements. Total blade area was estimated as blade length multiplied by mean width.

To enable measurements of blade elongation, two parallel holes were punched at the center of each blade 10 cm above the meristematic transition zone between blade and stipe, as described by Parke (1948). Elongation rates between sampling dates were then estimated based on the displacement of the holes from one sampling event to the next. After measuring and tagging, the specimens were returned to the sea in close proximity to their original position. As all individuals were naturally anchored by own holdfasts to smaller stones that could easily be collected together with each specimen, no artificial anchoring was needed. The site was revisited five times: May 17, 1999, July 7, 1999, September 6, 1999, November 8, 1999 and March 6, 2000. At these sampling events, the tagged specimens were measured as described above and new holes were punched. Between consecutive sampling events, there was a loss of 11–21 individuals and in May and November, the study was therefore supplemented with 24 and 16 new individuals, respectively. In July and September, the survey was expanded to include identical measurements at 4 and 11 m of depth as well. Forty individuals at each of the two depths were sampled for this purpose. Between each sampling date 12, 20–70 and 23 individuals were recovered at 4, 7 and 11 m of depth, respectively.

Density, biomass, morphometry and age determination

At each sampling date, additional individuals of S. latissima (7 m: n = 17–52, 4 and 11 m: n = 9–16) were collected for measurements of population structure (density and biomass per area), morphometry, area-specific DW, content of N and C in new and old blade tissue and determination of age. At 7 m of depth, samples were collected within 2–3 circles with a diameter of 2 m to assess the stock density. All individuals were harvested, counted and eventually dried at 60 °C to a constant weight for estimation of biomass per meter square. Due to low coverage at 4 and 11 m depth, density parameters were not rated here. Morphometric analysis included measurements of stipe length and diameter as well as age determination by counting growth rings in the stipe following the methodology of Parke (1948). Age determination was performed for 10–37 individuals per sampling and included only plants sampled at 7 m depth.

Area-specific dry weight and carbon and nitrogen contents

By the same interval as the measurements of blade width, three disks (3.2 cm in diameter) were cut across the blade to determine the area-specific DW after drying at 60 °C for 48 h. For each date and depth, the area-specific DW was determined for 10–20 individuals, dependent on the accessibility of plants. The area-specific DW was used to convert length—and area growth to biomass units. The disks were subsequently grinded and used for the determination of C and N contents. Carbon and Nitrogen contents were analyzed on an elemental analyzer (Roboprep C/N, Europa Scientific Ltd., UK) in line with a triple collector isotopic ratio mass spectrometer (Tracermass, Europa Scientific Ltd., UK). For each sampling date, five individuals from each of the three depths were analyzed for the content of C and N given as both tissue concentration (% of DW) and as area-specific contents (mg cm−2).

Large-scale patterns

Large-scale patterns in growth dynamics were assessed by comparing data on growth from the present study with literature findings at other latitudes, where growth was also typically assessed by consecutive blade-marking. The following studies covering the entire distribution range of S. latissima were included (latitudes in brackets): Fallis (1916) (48°N), Parke (1948) (56°N), Johnston et al. (1977) (56°N), Boden (1979) (42°N), Brady-Campbell et al. (1984) (41°N), Conolly and Drew (1985) (56°N), Dunton (1985) (70°N), Sjötun (1993) (60°N), Sjötun and Gunnarson (1995) (64°N), Lyngby and Mortensen (1996) (55°N), Schaeffelke et al. (1996) (54°N), Weile (1996) (55°N), Borum et al. (2002) (74°N), Andersen et al. (2011) (58°N) and Spurkland and Iken (2012) (60°N). Latitudes were provided directly by the cited literature or inferred from detailed descriptions of the study site. Data on average summer temperature were retrieved from the individual studies and included either as exact given values or as estimates based on graphs presented in the studies. For a few of the studies, data on depth were given as exact values as well.

Data analysis

Statistical analyses were performed with SAS-JMP 10.0 (SAS Institute Inc.,). The effect of latitude, temperature and depth on large-scale growth patterns (i.e., maximum growth rate and its timing) was initially assessed with general linear model (GLM) including depth and latitude or temperature as variables. As information of depth was only given for a few of the studies included, and it did not have any significant effect on any of the relations found (further elaboration in the results), depth was left out as an explanatory variable and the effects of latitude and temperature were subsequently assessed with linear regression.

Results are presented as mean ± SE if not stated otherwise. The effect of sampling depth was analyzed with ANOVA after confirming the assumptions of normality and homogeneity of variance. In case of significance of treatments, a post hoc test (Tukey HSD) was conducted. Data on age structure were analyzed using linear regression.

Annual turnover ratios were calculated on a length basis as well as on a DW basis. Both calculations were based on the total annual production in relation to a weighted annual average of the biomass.

Based on the blade-marking, new length growth was compared with total lengthwise changes of the blade, making it possible to estimate loss of tissue from the distal part. Net increase in DW, C and N contents as well as losses through abscission of old tissue were estimated between each sampling date by using the C and N contents and area-specific DW in new meristematic tissue and old (distal) tissue. Thus, within a given period, loss and growth were based on the average content of C and N at start and end of the period in, respectively, old and new tissue.

Temperature, salinity, irradiance and inorganic N concentration for the study site were presented as monthly means. For temperature and salinity, minimum and maximum values for each month were given as well. Inorganic N data included measurements of ammonium, nitrate and nitrite. Irradiance data at 7 m depth were obtained using average values of surface irradiance within day hours (i.e., values of 0, representing night hours, were omitted from the analysis) in combination with the light attenuation coefficient (Fig. 1).

Fig. 1
figure 1

a Seasonality of irradiance (µmol photons m−2 s−1) and inorganic nitrogen (µM) at the study site displayed as monthly averages. Irradiance data represent only averages of day hours and are calculated for 7 m depth based on surface irradiance and the light attenuation coefficient for the site at the different time points. Inorganic nitrogen data include levels of ammonium, nitrate and nitrite. b Seasonality of temperature (°C) and salinity (psu) at the study site displayed as monthly means. Vertical lines indicate minimum and maximum values within the months. Data are from the Danish monitoring program

Full size image

Results

Seasonality of physicochemical variables

The concentration of inorganic N followed an annual cycle typical for temperate areas with low levels during summer and autumn and higher levels in winter and spring months. From March to November, levels were lowest and did not exceed 1 µM inorganic N. During late autumn and winter, levels increased markedly to a peak of 14.4 µM inorganic N in January (Fig. 1a).

The annual average of Secchi depth was 6.1 m, monthly means ranging from 3.6 m in October to 8.9 m in September. Thus, irradiance showed considerable seasonal variation. Spring and summer irradiance at 7 m depth were in the range of 40–70 µmol photons m−2 s−1 with maximum levels in September, while levels between October and March were below 10 µmol photons m−2 s−1 (Fig. 1a). The drop in light intensity from September to October was a consequence of a massive phytoplankton bloom in early October. During summer months, when sampling was performed at different depths, the average daily light levels at 4, 7 and 11 m of depth were 128, 58 and 17 µmol photons m−2 s−1, respectively.

Monthly averages of salinity ranged between 19.8 psu in May and 26.7 psu in October. However, at least for shorter periods, lower values were reached as indicated by minimum values of 15.6 psu in May and several occurrences of salinities below 20 psu through spring and summer months (Fig. 1b). Monthly temperature averages ranged between 2.3 and 18.1 °C, with the lowest values recorded during winter and the highest during summer. Temperatures as high as 20.2 and 21.1 °C were measured in July and August, respectively (Fig. 1b).

Seasonal growth and loss dynamics

At 7 m depth, blade length and width, and thereby the total thallus area, increased through the spring and reached the maximum length of 128 ± 4.8 cm and width of 16 ± 1.1 cm in July (Fig. 2a, b). At this time of year, the lengths of the individuals at 11 and 7 m of depth were not significantly different from each other, whereas individuals at 4 m depth generally were smaller with an average length of 82.5 ± 8.2 cm (Table 1). From July to November, blade length decreased dramatically, with November values of 58 ± 5.7 cm at 7 m depth. The same pattern was found for the blade length of individuals at 4 and 11 m depth from July to September (Fig. 2a). Only minor changes were observed in blade length and width from November to March at 7 m depth. In comparison with length and width development (Fig. 2a, b), the seasonal pattern for the area-specific DW was slightly displaced with maximum values in September. However, combining the blade area and the area-specific DW to determine the seasonal variation of the biomass expressed as DW per individual, the maximum biomass was found in July (Fig. 2d), when the area was the largest even though the area-specific DW had not yet reached its maximum (Fig. 2c). In September, when area-specific DW peaked, no significant differences in area-specific DW were found between blades from 4 to 7 m depth, whereas blades from 11 m exhibited significantly lower values (Table 1).

Fig. 2
figure 2

Seasonal variations in a blade length, b blade width, c area-specific DW and d biomass pr. individual for S. latissima at different depths. Values are presented as mean ± SE. 4 m: n = 12–40; 7 m: n = 20–70; 11 m: n = 23–40. Blade length and width were measured on same individuals as used for growth measurement, whereas data on the area-specific DW were derived from individuals sampled for measurements of populations’ structure. Biomass per individual represents a combination of this information. DW dry weight

Full size image
Table 1 Maximum blade length and area-specific DW, daily growth and loss rates and tissue content of N and C for S. latissima from 4, 7 and 11 m of depth, based on data from July and September
Full size table

From March to July, daily growth in blade length, biomass and C and N contents in new meristematic tissue exceeded the daily loss through tissue abscission (Fig. 3a–d). Meristematic tissue exhibited the fastest rates of length growth (0.75 ± 0.04 cm day−1) (Fig. 3a) and N incorporation (2.7 mg N day−1 individual−1) (Fig. 3d) during spring (March–May), while the fastest daily increase in biomass (0.14 ± 0.012 g DW day−1 individual−1) (Fig. 3b) and C content (50 mg C day−1 individual−1) (Fig. 3c) occurred in early summer (May–July) at higher insolation (Fig. 1a). By contrast, from July to November, loss rates exceeded growth—and incorporation rates of all variables, with maximum loss rates during summer months, along with a reduction in growth and C and N incorporation. From November to March, rates of length growth and N incorporation in the new tissue exceeded loss rates, whereas growth and loss in terms of DW and C balanced each other (Fig. 3). Comparisons in July–September across the depth gradient showed a pattern of faster length growth but lower loss rates in deep compared to shallower water (Table 1).

Fig. 3
figure 3

Seasonal growth and loss dynamics of S. latissima expressed as average daily changes of a blade length, b dry weight (DW), c C content and d N content. Only data from 7 m depth are included. Values are presented as mean ±SE. n = 20–70. Data are obtained by combining growth measurements with average values of area-specific DW and C and N contents from the area-specific sampling. Growth represents new meristematic tissue, whereas loss represents distal tissue

Full size image

Nitrogen and carbon dynamics

The average N content of the blade (% of DW) showed distinct seasonal variation. It reached maximum in March (3.16 ± 0.12 and 3.48 ± 0.22 % of DW in March 1999 and 2000, respectively) and was reduced during summer to a minimum of 1.27 ± 0.04 % of DW in September (Fig. 4a). Area-specific N content did not show as pronounced seasonality, but overall, it increased from May to November and decreased during winter and spring (Fig. 4b).

Fig. 4
figure 4

Seasonal variation of N and C contents of S. latissima presented as a percentage of dry weight (DW) and b the area-specific content (mg cm−2). Each data point represents mean ±SE of five individuals, for which the average of 3–6 measurements along the blade length has been calculated

Full size image

The patterns of the C content were exactly the opposite. Carbon concentrations only varied slightly throughout the year with maximum levels of 37.7 ± 0.56 % of DW in November and minimum levels in March (34.8 ± 0.65 % of DW in 1999 and 22.7 ± 0.72 % of DW in 2000, Fig. 4a). However, during most of the study period, levels were in the range of 33–38 % of DW. In contrast, the area-specific content of C showed more marked variations throughout the year (Fig. 4b). Between May and November, the area-specific C content more than doubled from 2.43 ± 0.15 mg C cm−2 in May to 6.45 ± 0.57 mg C cm−2 in November, after which it decreased dramatically toward March 2000 to a minimum level of 1.65 ± 0.12 mg C cm−2. The seasonality of the C and N contents resulted in a peak atomic C/N ratio of 25 in September and minimum ratio of 9.4 and 5.6 in March 1999 and 2000, respectively.

Between July and September, the average tissue N concentration (% of DW) showed a significant increase toward deeper water, whereas the opposite was found for the C concentration (Table 1). The same tendencies occurred for the area-specific C and N contents (mg cm−2) (Table 1). At all depths, minimum values of N concentrations were found in September, amounting to 0.75 % of DW, 1.27 % of DW and 2.11 % of DW, respectively, for plants from 4, 7 and 11 m.

Population structure and production

Length and diameter of S. latissima stipes did not differ significantly over the season (ANOVA, P > 0.22, data not shown), and data from all sampling events were therefore pooled to analyze relationships between stipe dimensions and age (Fig. 5). The number of visible stipe growth rings varied from 1 to 3 for all collected specimens for the entire period of the survey, indicating that S. latissima in Aarhus Bay did not exceed the age of 3 years. There were significant correlations between stipe diameter and number of stipe growth rings (linear regression, r 2 = 0.27, F 1,110 = 40.44, P < 0.0001) as well as between stipe length and number of stipe growth rings (linear regression, r 2 = 0.14, F 1,110 = 18.04, P < 0.0001) (Fig. 5), suggesting that both parameters reflected the age of S. latissima. In addition, when blade length reached its maximum in July and loss through distal erosion was still limited, both stipe diameter and stipe length were significantly correlated to the length of the blade (Fig. 6), hence indicating that maximum blade length increased with age, at least until the age of three.

Fig. 5
figure 5

Diameter (a) and length (b) of S. latissima stipes as a function of the number of growth rings. Data are from March to November 1999 at 7 m of depth. n = 112. Statistics of the fitted lines: a r 2 = 0.27, F 1,110 = 40.44, P < 0.0001; b r 2 = 0.14, F 1,110 = 18.04, P < 0.0001

Full size image
Fig. 6
figure 6

Blade length of S. latissima as a function of stipe diameter and stipe length, respectively. Data are from July 1999, 7 m of depth. n = 9. Statistics of the fitted lines: stipe diameter, r 2 = 0.69, F 1,7 = 15.77, P = 0.0054; stipe length, r 2 = 0.77, F 1,7 = 22.84, P = 0.002

Full size image

Average annual length growth and blade length of S. latissima were, respectively, 132 and 84 cm, leading to a turnover ratio of 1.6 year−1 when calculated on a length basis. Likewise, the annual biomass turnover was 1.1 year−1 when calculated as the ratio between the average annual biomass production (17.8 g DW year−1) and weight per individual (16.2 g DW). The S. latissima population had an average density of 4.8 individuals m−2 on the scattered stones at 7 m depth and with an average weight of 16.2 g DW per individual, the area-specific biomass amounted to 77.5 g DW m−2 as an annual average. Maximum biomass of 122 g DW m−2 was found in July, whereas minimum biomass was found in March, being 24 g DW m−2 in March 2000 and 29 g DW m−2 in March 1999. Overall, the S. latissima population exhibited an annual blade production of 85.5 g DW m−2 at 7 m depth.

Large-scale patterns

Comparison of maximal daily elongation rates of this study (0.75 cm day−1 in March–May) with data covering the entire distribution range of S. latissima showed a fourfold variation across the distribution range with a general increase from north to south (Fig. 7a). On average, maximum daily growth rates increased 0.29 mm day−1 per latitudinal degree southwards. The comparison further indicated that the elongation rates of the Danish population studied here fell in the lower range of those registered at similar latitudes (Fig. 7a). The seasonal timing of the maximum elongation rate of the Danish kelp population fell within the range (March–May) identified for other populations within the latitude range of 45–60°N (Fig. 7c). By contrast, the maximum elongation rates in arctic regions occurred later in summer (June and July). Both the maximum elongation rate and its timing were correlated with water temperature (Fig. 7b, d), increasing the maximum growth rate with 0.47 mm and advancing it by 3.4 days/° temperature increase. The large-scale comparison of S. latissima growth patterns included studies conducted at different depths—the vast majority at depths of <10 m. Though testing was only allowed for a few of these studies of which the exact depth was given, there were no indications that differences in depth influenced any of the general relations found in Fig. 7a (GLM, F 1,9 = 1.47, P = 0.2556), Fig. 7b (GLM, F 1,8 = 0.62, P = 0.4521) or Fig. 7d (GLM, F 1,8 = 0.15, P = 0.7070).

Fig. 7
figure 7

Maximum length growth rate (a, b) and time of maximum growth rate (c, d) of S. latissima as a function of latitude (a, c) and average summer temperature (July–September) (b, d) across the entire geographical distribution of the species. Function of the fitted lines: a y = −0.29x + 27.2, b y = 0.47x + 5.96 d y = −0.11x + 6.19. Statistic of the fitted lines: a r 2 = 0.32, F 1,21 = 9.73, P = 0.0052; b r 2 = 0.30, F 1,12 = 5.22, P = 0.0414; d r 2 = 0.40, F 1,12 = 8.05, P = 0.015. No significant correlation between time of maximum growth rate and latitude was found; hence, no fitted line is displayed in c. Data from the present study (open squares) are compared with literature values (filled circles) where growth was also typically assessed based on consecutive blade-marking

Full size image

Comparing the course of growth curves from the present study and from northern and southern areas of the distribution range of S. latissima (based on the work of Brady-Campbell et al. 1984 and Dunton 1985) suggests additional differences between populations inhabiting the north and south (Fig. 8): (1) Onset of growth occurs earlier in southern populations, (2) growth is terminated slightly earlier in southern populations, (3) the growth period has a longer extend in southern populations, and (4) summer growth declines faster in southern populations (Fig. 8).

Fig. 8
figure 8

Blade elongation rate (mm day−1) throughout the year at three different latitudes. The three curves are based on the following studies: 41°N—Brady-Campbell et al. 1984, 56°N—present study, 70°N—Dunton 1985. Dashed lines indicate that data are obtained from different growth seasons

Full size image

Discussion

Growth patterns: global and local regulators

The latitudinal patterns identified in the global comparison (Fig. 7) suggest that large-scale factors are important in regulating growth dynamics in S. latissima. Water temperature and light (irradiance and photoperiod) are key candidates for inducing these latitudinal differences in seasonality and peaks of kelp growth (Kain 1989). The global comparison indicated that both the maximum growth rate and its timing are affected by temperature (Fig. 7b, d), thus stressing its importance as a large-scale parameter regulating growth dynamics in S. latissima. It is believed that critically high summer temperatures are also responsible for determining the species southern distribution boundary (van den Hoek 1982; Breeman 1988), and the fact that growth rates increase with latitude without any decline near the southern limit of distribution (Fig. 7a) indicates that the expansion is not prevented by growth of the sporophyte, but more likely coupled to the survival of earlier developmental stages (i.e., spores and gametophytes) and/or sporogenesis, which requires a period of minimum 4 weeks with short daylengths at 15 °C or lower to occur (Müller et al. 2009). The identified effects of temperature support earlier indications that S. latissima may expand northwards in a warmer future (Müller et al. 2009; Krause-Jensen et al. 2012). Furthermore, warming might entail earlier timing and probably also longer extent of the growth season in addition to increased peak production. In contrast, populations at the southern end of the distribution range may experience increased temperature stress and loss of biomass in late summer. A recent study on the seasonality of eelgrass populations along large-scale gradients also suggests earlier timing of the growth season with higher temperature, but no indications of longer growth season, increased peak biomass or production in warmer waters (Clausen et al. 2014).

Besides temperature, sea ice cover has also been identified as a major regulator of the growth of S. latissima in arctic regions (Krause-Jensen et al. 2012) and is together with long polar winters a factor known to confine kelp growth to a short period during summer where light conditions are sufficient (Dunton 1985; Borum et al. 2002). These factors most probably contribute to the later timing of maximum growth as well as the shorter growth period identified for arctic regions as compared to more temperate areas in this study (Figs. 7c, 8).

In addition to large-scale parameters, the huge variation in growth rate within narrow geographical ranges (55–56°N) (Fig. 7a) implies that growth dynamics are also influenced by parameters varying on a local scale, for instance high summer temperatures and suboptimal salinities. Marked reductions of growth at only 5° deviations from the optimum temperature range for S. latissima (10–15 °C) have been reported previously (Kain 1979; Bolton and Lüning 1982; Andersen et al. 2013). Thus, it is plausible that the productivity of S. latissima in Aarhus Bay was periodically inhibited during the study period by non-optimal temperatures at the extreme of the recorded range from 2.4 °C (minimum value of March 1999) to 21.1 °C (maximum value of August 1999). It seems that high summer temperatures were more detrimental in the regulation of productivity than low winter temperatures, considering that the lowest growth rates and the highest loss rates coincided with high temperatures during summer, whereas a similar relationship was not apparent during cold winter months, where loss rates were low (Figs. 1b, 3a–b). This pattern is in accordance with previous findings that high temperatures (20 °C) lead to reduced net photosynthesis along with increased tissue deterioration as compared to lower temperatures (10 and 15 °C) (Andersen et al. 2013). Krumhansl and Scheibling (2011) also found this effect of temperature but further pinpointed the coverage of epiphytes as an enhancer of blade tissue loss. For this present study, however, no notable epiphytes were seen and hence were not considered problematic in that term.

Low and fluctuating salinity, typical for inner Danish waters, may also have contributed to the relatively low production, as osmotic stress can exert physiological stress on kelps. During this study, monthly averages of salinity ranged between 19.8 and 26.7 psu, which are somewhat below the optimal range for S. latissima of 27–33 psu reported by Gerard et al. (1987). Periodically, salinities were as low as 15.6 psu, which may have affected the growth negatively and contributed to the low production in the area. Low growth of S. latissima has also been found in another Danish study (5.4 mm day−1, Weile 1996) conducted in an area of low salinity (annual average <14 psu, data obtained from DNAMAP) reinforcing the linkage of low growth and low salinity.

Annual production and turnover

In average, S. latissima exhibited an annual length growth of 132 cm at 7 m depth. A global comparison of the annual length growth reveals a tendency of increasing values in a north to south direction (Krause-Jensen et al. 2012) as was also found for maximum growth rates in this work. Within this global data comparison of annual length growth, Danish data once again fall in the lower range.

Based on the blade length, the annual turnover amounted to 1.6 year−1 on average, whereas the turnover based on DW was somewhat lower, 1.1 year−1, stressing that the correlation between length and weight is not linear and that as an individual grows longer, it also gets wider and thicker. Turnover rates are prone to vary with age (Mann 1972), but since determinations of age and blade elongation (i.e., production) were not made for the same individuals in the present study, an exact comparison of turnover values between age classes was not possible. However, we found that blade length increased with age (Figs. 5, 6) and assuming that absolute length growth was similar in all age classes (Mann 1972), data indicated a reduced turnover in older individuals. Our results thus support previous statements concerning the influence of cohort distribution within a population on growth dynamics of Laminariales (Parke 1948; Mann 1972; Lüning 1979; Sjötun 1993).

Standing stock and area production

The average standing stock of S. latissima in Aarhus Bay (77.5 g DW m−2) was low compared to kelp biomasses reported from elsewhere (Bartsch et al. 2008). A major reason for this was a limitation of suitable substratum. The scattered stones of the study site only allowed an average density of 4.8 individuals m−2, whereas rocky shores can host much denser populations (Mann 1972). In addition, the size of the individuals (weighted length average: 83.8 cm; weighted biomass average: 16.1 g DW individual−1) was also relatively modest compared to specimens found elsewhere (i.e., Parke 1948; Gaevert et al. 2001; Krause-Jensen et al. 2012) and thus contributed to the low standing stock in the area. It is probable that the relatively small stones which composed the substratum could not support larger individuals. The limited standing stock along with the moderate turnover ratio implied that annual area production (85.5 g DW m−2) was also modest relative to levels realized for kelp beds elsewhere (Mann 1973; Lüning 1990; Vadas et al. 2004).

Carbon and nitrogen dynamics

Generally, the lamina contained more C during summer and autumn than during spring months, corresponding to the suggestion that S. latissima stores carbohydrates during summer for use in winter growth (Black 1950). Also, the increase in blade length from November to March (Fig. 3a) simultaneously with a decrease in the C content (Fig. 4a, b) indicated a consumption of stored carbohydrates. These findings are comparable to what has been described for Laminariales elsewhere (Mann 1973; Sjötun 1993). However, our study also showed that considerable amounts of the C were lost through abscission of distal tissue during summer months (Fig. 3c), hence proposing that the ability to translocate C from old tissue to new meristematic tissue was not complete.

The tissue C content decreased with increasing depth paralleling the decline in irradiance (Table 1), merely indicating that the ability to store carbohydrates was influenced by the light conditions, thus again stressing the important role of light for the biomass dynamics of S. latissima. Depth also influenced loss rate (Table 1); the highest loss rates were found at 4 m depth presumably as a consequence of a higher degree of exposure to pressures—in terms of waves, low salinity and high temperatures. Despite a higher C content in these shallow growing algae, the high loss of distal blade tissue reduces the translocation of stored C for new meristematic growth and consequently, this might along with the higher exposure to pressures reduce their growth performance.

From September to March, the total N content in S. latissima increased as indicated by a dramatic increase in tissue N concentration as well as an overall increase in the area-specific N content (Fig. 4a, b). This increase coincided with the period of maximum ambient nutrient levels (Fig. 1b) and verified the ability of the species to build up N reserves under such conditions. On the contrary, between March and September, nutrient levels were low in Aarhus Bay (Fig. 1b). At this time of the year, the N tissue concentration dropped dramatically (Fig. 4a) simultaneously with an occurrence of marked growth in blade area and biomass (Fig. 2a, b). Hence, the decline of tissue N concentration was a consequence of N being diluted as new tissue was formed. In other words, internally stored N was utilized to support new growth in this period. However, as with C, considerable amounts of N were lost from the distal tissue during summer (Fig. 3d) and the ability of S. latissima to recover N from the distal tissue was not complete. High growth presumably based on internal N reserves was only maintained until July, after which it ceased (Fig. 3a, b). In this same period, the internal N concentration decreased to 1.3 % of DW, which is below the critical N value (1.7 % of DW) necessary to maintain maximum growth reported for other brown seaweeds (Pedersen and Borum 1996). It therefore seems reasonable to assume that growth decline during summer was partly related to N limitation, as ambient nutrient levels at this time a year were low (Fig. 1a) and internal storages were exhausted (Fig. 4a).

At 11 m of depth, the minimum tissue N concentration was 2.1 % of DW, somewhat higher than the critical N value (Pedersen and Borum 1996) and hence, the same exhaustion of N did not seem to occur in deeper waters despite higher elongation rates. The presence of a pycnocline between 7 and 11 m of depth during summer months may have caused this pattern by enhancing nutrient availability, and hence nutrient uptake, below the pycnocline as compared to above it. Further expected benefits of a summer pycnocline in terms of increased salinity and reduced temperature did not seem to enhance the overall growth performance of the plants at 11 m depth, as even though they had a higher elongation rate, their low area-specific DW (Table 1) merely indicated an increase in the blade area rather than a real growth in biomass. Increasing the blade area could be an adaptation to better light exploitation in this low light environment (17 µmol photons m−2 s−1).

In summary, our study confirms earlier findings that seasonal growth and loss dynamics of S. latissima are indeed regulated by annual fluctuations in light and N levels. However, differences in light and exposure along depth gradients also affect the dynamics, and high losses of both N and C through summer abscission of distal tissue question the efficiency of translocation in S. latissima, especially where late summer losses are accelerated by, e.g., high temperature. Our large-scale comparisons further highlight that warming advances the timing and increases the level of peak growth and also seems to expand the growth period. It is suggested that local factors such as low salinity and periodically high summer temperature may contribute to explaining why maximum growth rate in the present study falls in the low range of values measured in other S. latissima populations from similar latitudes.